Adsorption technologies for the recycling of fluoride. Hydrometallurgical remediation of spent potlining and hypercrosslinked polymeric extractants

Fluorite (CaF2) is a critical natural reserve and industrial fluoride recycling is of global importance. This work approaches the problem on two fronts. First, modification of a hydrometallurgical treatment system for spent potlining (SPL); a fluoride-rich, toxic, solid aluminium industry waste. Second, fluoride adsorption by metallated, porous, hydrophilic hypercrosslinked polymers (HHCPs). To ameliorate hydrometallurgical SPL treatment, an adsorption system is investigated, to selectively extract fluoride from a mixed leachate stream. A lanthanum-loaded chelating resin (La-MTS9501) is chosen for this purpose and its uptake characteristics are investigated. La-MTS9501 co-extracts fluoride and aluminium from simulated SPL leachate via adsorption of aqueous aluminium hydroxyfluoride (AHF) complexes, which chelate to the La centres via bridging fluoride or oxide ligands, with additional weaker interactions at higher leachate concentrations. Uptake was described best by multilayer and heterogeneous isotherm models. The theoretical maximum fluoride uptake capacity (qmax) was 145 mg·g-1 and the uptake followed pseudo second-order kinetics, with film-diffusion likely being rate-controlling. In dynamic column simulations, fluoride breakthrough was well-described by the Dose-Response model and dynamic loading capacity was 66.7 mg·g-1. Real SPL samples were leached via a two-step treatment using 1M NaOH with added H2O2, then 0.5 M H2SO4, which solubilised the majority fluoride content of numerous samples, of different mineral composition. The good dynamic column performance achieved with a simulant feed was retained with real SPL leachate. The fluoride and aluminium were selectively eluted from the column, using 1 M NaOH, with few problematic cocontaminants. Synthetic cryolite could potentially be attained from this feed by precipitation. A first-order technicoeconomic model suggested that, with some process optimisation, the enhanced hydrometallurgical system could offset its fixed-CAPEX within six years. Microporous hydrophilic hypercrosslinked polymers termed “HHCP1” and “HHCP2” were synthesised respectively from 2,2’-biphenol and bisphenol A monomers. These materials had high exchange capacities of up to 6.34 mmol·g-1 and unexpectedly broad pKa ranges. Polymers were loaded with calcium, to create fluoride selectivity, achieved by covalent bonding and unexpectedly by the formation of CaCO3, from adsorbed atmospheric CO2. Fluoride was extracted by conversion of the bound Ca species to CaF2. HHCP1 retained a high surface area after metallation and had a theoretical qmax of 267 mg·g-1. HHCP2 was converted to a surface-functionalised material and had lesser capacity, but faster uptake kinetics. Both polymers could be used in a dynamic column system and were selective in mixed anions uptake experiments, although phosphate was co-extracted. They could also be successfully regenerated in a two-step treatment, using HNO3, then re-metallation, over four cycles of use.